Factor, Nested

General Overview

Two factors are said to be “crossed” when every category of one factor overlaps every category of another. Consider Figure 1, a crossed design. This figure visually represents the design for a study in which participants are expected to work with a confederate who will speak from Script 1 or Script 2 [factor = type of script (script 1, script 2)]. In addition, in this study, some participants are caffeinated, and others are not [factor = caffeination status (not caffeinated, caffeinated)]. This is a crossed design because every possible combination of factors is tested and measured.

Figure 1 Crossed Design

Now consider Figure 2, a nested design. A factor is said to be nested within another factor when each category of the first factor does not crossover with every category of the second factor. Consider the same scenario of a study in which the confederate is trained to interact with participants using different scripts and only half of the participants are caffeinated before the experiment. The design, visually represented in Figure 2, shows that caffeinated participants will never interact with a confederate using Script 3 or 4, nor will a noncaffeinated participant interact with a confederate who is using Script 1 or 2. In other words, depending on caffeination status, a participant will be exposed to only select levels of the other factor, type of script. Thus, the scripts are nested within the caffeine factor.

Figure 2 Nested Design

[Page 497]Using a nested design greatly limits the knowledge gained by using a factorial design (i.e., a design that contains more than one factor) in that it does not produce an interaction effect (i.e., a change in the main effect of one factor over the levels of the second factor). Yet, sometimes a nested design is more practical for the needs of the researcher. Consider the following example: